Core ML: Integrate machine learning models into your app

https://developer.apple.com/documentation/coreml

by yonilevy • 9 years ago

171 40 9 years ago

developer.apple.com

eggie54• 9 years ago

Apple's new iOS CoreML inference engine supports Keras models! Developers will be able to design and train model using Keras and then convert the architecture to run on the CoreML engine. I suppose you can run TensorFlow models too if you designed them w/ Keras.

likelynew• 9 years ago

Yeah, that is surprising given that keras is not a language, but a continuously changing python class. How bug complete is the conversion? What about some new features like attention? And what about the future changes in keras?

jordansmithnz• 9 years ago

I think they mentioned in the Core ML lab that the conversion tool is being open sourced - which would help with these issues.

kensai• 9 years ago

"Core ML is optimized for on-device performance, which minimizes memory footprint and power consumption."

This is major, if they have managed to achieve it reasonably. But before opening a Sekt, I want to see some benchmarks. :)

singularity20012• 9 years ago

"BNNS does not do training, however. Its purpose is to provide very high performance inference on already trained neural networks." :(

https://developer.apple.com/documentation/accelerate/bnns

eggie54• 9 years ago

yes, it's an inference engine that runs pre-built architectures. why would you want to train on the device?

singularity20012• 9 years ago

Maybe not fully train but fine tune on the user data ( images, voice, surroundings, etc, etc)

cshenton• 9 years ago

Federated Learning. It's something Google does at the Android level, but I don't think they provide an API for app developers to train on device.

Another use case is privacy: ship a pretrained facenet style model and let it learn to map your address book to faces in your photos without your data leaving the device.

igravious• 9 years ago

BNNS = “Basic neural network subroutines” if you are not interested in clicking through!

seanmcdirmid2• 9 years ago

When "Core ML" is no longer a mini language to reason about ML the PL semantics...

pjmlp• 9 years ago

Yeah, it crossed my mind as well.

But nowadays every time I see ML on HN, it happens to mean ML = Machine Learning instead of ML = programming language.

simplify• 9 years ago

It's also going to make the name ReasonML even more confusing, considering you can apply the word "Reason" to machine learning...

killin_dan• 9 years ago

Machine Learning needs a new acronym. ML is already taken! Get your own!

castis• 9 years ago

Which of these 109 entries is objectively the definitive one?

http://www.acronymfinder.com/ML.html

Skunkleton• 9 years ago

At this point, I would like to nominate Malt Liquor.

hackerbot• 9 years ago

using it means need to handle Android with another framework separately

mksmurf7772• 9 years ago

Is there any cross-platform framework recommended?

asadlionpk• 9 years ago

check Caffe 2. I haven't tried that yet but looks promising.

mtw• 9 years ago

Why not caffe2? And why mention libsvm, are we in the 90s?

wodenokoto2• 9 years ago

Because SVM is fast and more than good enough for most problems.

eggie54• 9 years ago

I think he's talking about libsvm, the c++ package

ska• 9 years ago

   And why mention libsvm, are we in the 90s.

This is quite an odd question. It's not like CNNs entirely cover the same problem domain well. So why wouldn't you want SVM support?

blt• 9 years ago

Cool that the file format is officially specified as a Protocol Buffer: https://developer.apple.com/documentation/coreml/converting_... (bottom of page)

simplehuman• 9 years ago

Is there a protobuf to capnproto converter?

jamesswift3• 9 years ago

Federighi says Core ML on iPhone is 6x faster than Google Pixel and Samsung Galaxy S8.

How they actually compare?

suyash• 9 years ago

The probably ran the same tests on Android using Pixel and S8 and then on iPhone (which leveraged CoreML). I would like to see a more detailed analysis myself.

db3d• 9 years ago

Lots of options to explore, but no reinforcement learning yet.

Also, some converted Core ML Models ready to use here: developer.apple.com/machine-learning

halflings• 9 years ago

Reinforcement Learning relates to the way you train your model. Most of the time, it ends up being a feed-forward neural net (possibly with some convolutional layers), and rarely an RNN, all of which are supported.

ajay-d• 9 years ago

The python package, coremltools, to convert the trained model is only for python 2.7??

breatheoften• 9 years ago

Is python3 shipped on OS X by default? I'm not sure why they are still shipping 2.x -- anyone know if 3.x will be shipped in high sierra?

A bit disappointed that the model conversion tool only supports an older version of Keras as well (1.2.2). Keras 2.0 is pretty new but I hope they update the conversion tools for it quickly ...

I wonder if the conversion tool will be open source ... seems like they'd want to support the widest net of external models since they don't yet have a way to produce .coreml models directly. Or maybe the intent is to augment Keras/caffe/etc to support saving .coreml directly?

saagarjha• 9 years ago

macOS comes with Python 2 by default.

jdr0317• 9 years ago

Both python 2 (/usr/bin/python) and python 3 (/usr/bin/python3) are shipped by default on Ubuntu.

Right now, coremltools is only available for Python 2.7 (https://pypi.python.org/pypi/coremltools), which is annoying as the entire code base I've worked on for months at my current firm is in Python 3.6. Hopefully this is updated soon for Python 3 support.

OldeElk• 9 years ago

Does it support caffe/TensorFlow/MXnet model inferencing?

dmix• 9 years ago

Having a directory of trained models to download is an interesting concept. This will certainly accelerate the adoption of ML.

Core ML: Integrate machine learning models into your app

Related Stories